Victor Perez
Victor Perez
> Question: why do we need to clone the kernel and device functions into a new module ? I do not understand the motivation for this PR fully. I am...
Hi, @wsmoses, We have this in our task list, but, as this is not in our critical path, we will wait for now. However, we have discussed this and agreed...
> @victor-eds please, attach the code. Added inline
> Some massaging of our `enable_if`s is required here: > > https://github.com/intel/llvm/blob/5fae0aaa4f932c6ecddb052a0cb3d873ff2eee64/sycl/include/sycl/kernel_handler.hpp#L66-L85 > > We have a simple switch between `compound` and `fundamental` types, but we should probably use less...
Fixed by #12925
@v-klochkov proposed an additional variant of `private_alloca` receiving a required alignment. This would be a great addition to the current API. I was thinking on something like: ```c++ namespace sycl::ext::oneapi::experimental...
> > @v-klochkov proposed an additional variant of `private_alloca` receiving a required alignment. This would be a great addition to the current API. I was thinking on something like: >...
@v-klochkov I added the second signature. Feel free to comment.
@gmlueck @v-klochkov This PR has been approved by the required reviewers, but, as you also provided feedback, can you please confirm current status looks good to you? Please, resolve the...
> @victor-eds, could you change the file name from `sycl_ext_private_alloca.asciidoc` to `sycl_ext_oneapi_private_alloca.asciidoc`, please? Done (#12881). Thanks for pointing out!